Design Arena: game dev leaderboard
- GPT-6 Astra leads
- 1474
- models measured
- 127
- lead over #2
- +5 Elo
- median
- 1195
Leader on Design Arena: game dev · read October 21
On October 21, 2026, 127 of our catalog's models had a Design Arena: game dev result, and OpenAI GPT-6 Astra led with 1474.
Design Arena shows people two designs made by two models from the same prompt (a small game) and asks which is better. The number is the Elo rating built from those votes, as Design Arena publishes it.
Role in the index
We collect this benchmark and show it as supporting evidence. It is not part of the Vibe Coding Index, because it measures something other than coding and the index is about coding. Read the methodology
| # | Model | Result | Battles | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|
| 1 | GPT-6 AstraOpenAI | 1474 | 660 | 88.7 | $50.00 | |
| 2 | Claude Sonnet 5.5Anthropic | 1469 | 563 | 88.9 | $10.00 | |
| 3 | Claude Opus 5.5Anthropic | 1457 | 782 | 92.2 | $20.00 | |
| 4 | GPT-6.1 SolOpenAI | 1389 | 728 | 88.7 | $10.00 | |
| 5 | Claude Fable 5.1Anthropic | 1382 | 1,497 | 89.1 | $50.00 | |
| 6 | Muse WonderMeta | 1380 | 777 | 64.5 | $20.00 | |
| 7 | Gemini 4 ArgonGoogle | 1377 | 1,261 | 80.5 | $10.00 | |
| 8 | GPT-5.6 SolOpenAI | 1368 | 2,238 | 76.7 | $20.00 | |
| 9 | DeepSeek V4.1 Flash 1017DeepSeek | 1365 | 1,045 | 83.3 | $0.60 | |
| 10 | Kimi K3Moonshot AI | 1361 | 847 | 61.8 | $11.936 | |
| 11 | Claude Opus 5Anthropic | 1339 | 3,051 | 81.8 | $25.00 | |
| 12 | Claude Fable 5Anthropic | 1336 | 3,461 | 82.1 | $50.00 | |
| 13 | DeepSeek V4.1 FlashDeepSeek | 1331 | 12,255 | 77.2 | $0.525 | |
| 14 | Muse Spark 1.3Meta | 1329 | 12,087 | 79.7 | $4.25 | |
| 15 | Gemini 3.7 FlashGoogle | 1318 | 14,133 | 69.9 | $3.75 | |
| 16 | Claude Haiku 5.5Anthropic | 1315 | 1,605 | 75.9 | $0.50 | |
| 17 | Grok 4.7xAI | 1314 | 703 | 75.0 | $6.00 | |
| 18 | MiMo-V2.6-ProXiaomi | 1314 | Not published | 72.6 | $0.87 | |
| 19 | GPT-6 SolOpenAI | 1313 | 1,356 | 82.8 | $10.00 | |
| 20 | Qwen3.8 Max (0902)Qwen | 1313 | 2,648 | 63.5 | $6.00 | |
| 21 | GLM 5.3Z.ai | 1311 | Not published | 73.1 | $1.40 | |
| 22 | Gemini 3.8 FlashGoogle | 1305 | 12,533 | 70.5 | $4.125 | |
| 23 | Muse Spark 1.2Meta | 1302 | 12,865 | 65.1 | $4.25 | |
| 24 | Grok 4.6xAI | 1296 | Not published | 76.4 | $6.00 | |
| 25 | GPT-5.5OpenAI | 1294 | 7,210 | 67.8 | $30.00 | |
| 26 | Muse Spark OpenMeta | 1290 | 204 | 62.1 | $4.25 | |
| 27 | GLM 5.3 FlashZ.ai | 1279 | Not published | 71.0 | $0.50 | |
| 28 | Claude Opus 4.6Anthropic | 1278 | 2,745 | 66.9 | $25.00 | |
| 29 | Claude Sonnet 5Anthropic | 1278 | 3,482 | 63.5 | $10.00 | |
| 30 | Muse Spark 1.1Meta | 1269 | Not published | 48.9 | $4.25 | |
| 31 | GLM 5.2Z.ai | 1268 | 7,859 | 57.7 | $7.00 | |
| 32 | Qwen3.7 MaxQwen | 1265 | 6,527 | 45.1 | $4.425 | |
| 33 | GLM 5.1Z.ai | 1265 | Not published | 44.0 | $4.40 | |
| 34 | Claude Opus 4.8Anthropic | 1264 | 4,936 | 66.0 | $25.00 | |
| 35 | Gemini 3.5 FlashGoogle | 1263 | 6,148 | 42.8 | $9.00 | |
| 36 | Gemini 3.6 FlashGoogle | 1263 | 5,663 | 48.3 | $3.75 | |
| 37 | Grok 4.5xAI | 1263 | Not published | 75.5 | $6.00 | |
| 38 | Qwen3.7 PlusQwen | 1262 | 3,669 | Not benchmarked | $1.28 | |
| 39 | MiMo-V2.5-ProXiaomi | 1261 | 4,821 | 40.0 | $0.609 | |
| 40 | Claude Sonnet 4.6Anthropic | 1259 | 4,485 | 57.8 | $15.00 | |
| 41 | GPT-6 LunaOpenAI | 1259 | 1,395 | 71.4 | $0.50 | |
| 42 | Kimi K2.6Moonshot AI | 1251 | 5,031 | 44.2 | $3.41 | |
| 43 | MiMo-V2.5Xiaomi | 1249 | 5,729 | 34.0 | $0.238 | |
| 44 | GLM 5 TurboZ.ai | 1249 | 5,414 | Not benchmarked | $4.00 | |
| 45 | GLM 5Z.ai | 1234 | 5,689 | 59.7 | $3.20 | |
| 46 | GPT-5.6 TerraOpenAI | 1230 | 14,843 | 69.0 | $12.00 | |
| 47 | Claude Opus 4.5Anthropic | 1229 | 5,034 | 66.8 | $25.00 | |
| 48 | DeepSeek V4 Pro 0423DeepSeek | 1229 | 4,413 | 35.3 | $1.9105 | |
| 49 | Mistral Large 4Mistral AI | 1226 | 1,406 | 55.3 | $4.18 | |
| 50 | R Beam 1.0 501B A23BReflection | 1222 | Not published | 33.8 | $1.00 | |
| 51 | Kimi K2.7 CodeMoonshot AI | 1220 | 3,730 | 39.1 | $3.35 | |
| 52 | GLM 5V TurboZ.ai | 1219 | 7,898 | 28.8 | $4.00 | |
| 53 | Qwen3.6 PlusQwen | 1211 | 8,199 | 37.4 | $1.95 | |
| 54 | Mercury 3Inception | 1210 | 413 | Not benchmarked | $0.35 | |
| 55 | MiniMax M2.7MiniMax | 1210 | Not published | 28.4 | $1.20 | |
| 56 | MiniMax M3MiniMax | 1210 | 4,461 | 57.8 | $1.20 | |
| 57 | DeepSeek V4 Flash 0731DeepSeek | 1208 | 1,761 | Not benchmarked | $1.32 | |
| 58 | Hy3Tencent | 1208 | 3,123 | 43.9 | $0.528 | |
| 59 | Kimi K2.5Moonshot AI | 1207 | 5,004 | 58.5 | $2.50 | |
| 60 | Gemini 3.1 Pro PreviewGoogle | 1203 | 7,075 | 35.3 | $12.00 | |
| 61 | DeepSeek V4 Flash 0423DeepSeek | 1197 | 4,371 | 54.6 | $0.182 | |
| 62 | GPT-5.2OpenAI | 1196 | Not published | 58.9 | $14.00 | |
| 63 | GPT-5.6 LunaOpenAI | 1196 | 3,914 | 61.5 | $1.20 | |
| 64 | GPT-5.4OpenAI | 1195 | 4,000 | 37.7 | $15.00 | |
| 65 | Grok 4.20xAI | 1192 | Not published | Not benchmarked | $2.50 | |
| 66 | GLM 4.7Z.ai | 1189 | 5,108 | 33.6 | $1.98 | |
| 67 | GPT-5OpenAI | 1185 | 1,283 | 72.2 | $10.00 | |
| 68 | GPT-5.1OpenAI | 1179 | 1,812 | 53.3 | $10.00 | |
| 69 | MiniMax M2.5MiniMax | 1176 | Not published | 59.4 | $0.95 | |
| 70 | Grok 4.3xAI | 1174 | 7,365 | 22.5 | $2.50 | |
| 71 | Claude Opus 4.1Anthropic | 1172 | 1,524 | 27.2 | $75.00 | |
| 72 | Solar Pro 4Upstage | 1170 | 1,025 | 24.4 | $0.60 | |
| 73 | Claude Sonnet 4.5Anthropic | 1166 | 6,024 | 59.1 | $15.00 | |
| 74 | Gemini 3 Flash PreviewGoogle | 1166 | 369 | 75.8 | $3.00 | |
| 75 | T InklingThinking Machines | 1166 | 3,564 | 49.7 | $4.05 | |
| 76 | GPT-5.3-CodexOpenAI | 1162 | 2,963 | 29.5 | $14.00 | |
| 77 | Step 3.7 FlashStepFun | 1158 | 4,539 | Not benchmarked | $1.15 | |
| 78 | GLM 4.6Z.ai | 1151 | 1,929 | 52.2 | $2.20 | |
| 79 | DeepSeek V3.2 ExpDeepSeek | 1145 | 2,056 | 55.0 | $0.41 | |
| 80 | GLM 4.5Z.ai | 1145 | 1,996 | 64.2 | $2.20 | |
| 81 | Qwen3.5 397B A17BQwen | 1141 | 1,162 | 28.6 | $2.34 | |
| 82 | Nemotron 3 UltraNVIDIA | 1138 | 4,982 | Not benchmarked | $2.40 | |
| 83 | GPT-5.1-CodexOpenAI | 1138 | 182 | 50.5 | $10.00 | |
| 84 | MiniMax M2.1MiniMax | 1135 | 2,300 | Not benchmarked | $1.20 | |
| 85 | GLM 4.7 FlashZ.ai | 1135 | 1,560 | Not benchmarked | $0.40 | |
| 86 | DeepSeek V3.2DeepSeek | 1132 | 4,509 | 54.4 | $0.42 | |
| 87 | DeepSeek V3.1 TerminusDeepSeek | 1130 | Not published | Not benchmarked | $1.00 | |
| 88 | GPT-5 MiniOpenAI | 1130 | 5,246 | 59.8 | $2.00 | |
| 89 | MiniMax M2MiniMax | 1119 | Not published | 45.3 | $1.02 | |
| 90 | Gemini 2.5 ProGoogle | 1112 | 633 | 37.0 | $10.00 | |
| 91 | Solar Mini 4Upstage | 1110 | 445 | Not benchmarked | $0.20 | |
| 92 | Nemotron 3.5 Super VLNVIDIA | 1108 | 960 | 65.5 | $0.50 | |
| 93 | Qwen3.5 Plus 2026-02-15Qwen | 1106 | 2,804 | Not benchmarked | $1.56 | |
| 94 | Qwen3 Coder 480B A35BQwen | 1099 | 155 | 50.0 | $1.00 | |
| 95 | R1 0528DeepSeek | 1098 | 1,636 | 71.4 | $2.15 | |
| 96 | Claude Haiku 4.5Anthropic | 1097 | 4,502 | 49.8 | $5.00 | |
| 97 | GLM 4.5 AirZ.ai | 1094 | 1,660 | Not benchmarked | $0.85 | |
| 98 | GPT-5.1-Codex-MiniOpenAI | 1093 | 4,185 | 6.2 | $2.00 | |
| 99 | DeepSeek V3.1DeepSeek | 1085 | Not published | Not benchmarked | $0.95 | |
| 100 | Mistral Large 3 2512Mistral AI | 1080 | 4,125 | Not benchmarked | $1.50 | |
| 101 | GPT-4.1OpenAI | 1080 | 181 | 46.5 | $8.00 | |
| 102 | Trinity Large ThinkingArcee AI | 1079 | 2,073 | 5.4 | $0.80 | |
| 103 | Mistral Medium 3.5Mistral AI | 1079 | 1,521 | 9.2 | $7.50 | |
| 104 | Mistral Medium 3.1Mistral AI | 1074 | 3,816 | Not benchmarked | $2.00 | |
| 105 | GPT-4.1 MiniOpenAI | 1071 | 224 | 28.5 | $1.60 | |
| 106 | DeepSeek V3DeepSeek | 1054 | 1,891 | 70.2 | $1.0287 | |
| 107 | GPT-5 NanoOpenAI | 1045 | 533 | 34.8 | $0.40 | |
| 108 | Ministral 3 14B 2512Mistral AI | 1038 | 149 | Not benchmarked | $0.20 | |
| 109 | o3OpenAI | 1034 | Not published | 70.7 | $8.00 | |
| 110 | Gemini 3.1 Flash Lite PreviewGoogle | 1029 | 3,706 | 8.1 | $1.50 | |
| 111 | Mistral Medium 3Mistral AI | 1018 | 599 | Not benchmarked | $2.00 | |
| 112 | Gemini 2.5 FlashGoogle | 1005 | 514 | 28.7 | $2.50 | |
| 113 | o4 MiniOpenAI | 1004 | Not published | 68.6 | $4.40 | |
| 114 | gpt-oss-120bOpenAI | 992 | 445 | 34.5 | $0.60 | |
| 115 | Ministral 3 8B 2512Mistral AI | 990 | 168 | Not benchmarked | $0.15 | |
| 116 | Mercury 2Inception | 985 | 4,656 | 0.0 | $0.75 | |
| 117 | Kimi K2 0711Moonshot AI | 974 | 138 | 64.7 | $2.30 | |
| 118 | GPT-4.1 NanoOpenAI | 971 | 250 | 8.9 | $0.40 | |
| 119 | Codestral 2508Mistral AI | 968 | 576 | Not benchmarked | $0.90 | |
| 120 | Qwen3 235B A22B Thinking 2507Qwen | 957 | 514 | Not benchmarked | $3.50 | |
| 121 | Qwen3 235B A22B Instruct 2507Qwen | 952 | 558 | Not benchmarked | $0.55 | |
| 122 | Ministral 3 3B 2512Mistral AI | 949 | 230 | Not benchmarked | $0.10 | |
| 123 | GPT-4oOpenAI | 906 | 241 | 38.8 | $10.00 | |
| 124 | Qwen3 30B A3BQwen | 899 | 224 | Not benchmarked | $0.50 | |
| 125 | Mistral Small 3.2 24BMistral AI | 888 | Not published | Not benchmarked | $0.25 | |
| 126 | Llama 4 MaverickMeta Llama | 838 | 205 | 18.1 | $0.6525 | |
| 127 | Llama 4 ScoutMeta Llama | 773 | 175 | 9.1 | $0.30 |
Only models in our catalog are listed, and each appears once with its best published run. 141 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.